briven · docs

roadmap

what we're building, in rough order of priority. dates are intent, not commitment — anything ships when it's ready and not before. the changelog is what actually landed; this page is what we're working towards.

now

dogfooding briven on a single VPS, hardening the operator surface, public alpha behind invite codes.

  • [done]phase 0 — foundations: api, runtime, realtime, dashboard, docs all live
  • [done]phase 1 — reactive queries via LISTEN/NOTIFY across all client SDKs
  • [done]phase 2 — per-tier rate limits, abuse pipeline, audit log, encryption-at-rest for env vars
  • [done]phase 3 — usage metering, beta invite UX, studio CRUD UI, admin triage
  • [done]phase 4 — self-host templates, observability stack, public docs polish, status page
  • [now ]deploy on briven's own VPS, run alpha invitees through the cli + dashboard

next

closing the gaps that turn alpha into a product anyone can adopt without hand-holding.

  • [next]point-in-time recovery (replace nightly pg_dump with WAL streaming + 7-day window)
  • [next]multi-region read replicas (eu-west + us-east, opt-in per project)
  • [next]team auth: SSO via SAML/OIDC, per-project member roles beyond owner/admin/developer
  • [next]payments live: Polar + invoice PDFs + dunning + tax handling for EU/UK
  • [next]function logs: full-text search across the last 7 days (today is tail-only)
  • [next]scheduled functions: cron expressions in code, observable in the dashboard
  • [next]file uploads: presigned-URL flow + per-project minio bucket + image transforms

later

quality-of-life and breadth. nothing here blocks GA but we know we'll want them.

  • [next]first-class python + go SDKs (parity with TS/JS clients)
  • [next]vector search beyond pgvector — first-party embedding generation via the platform
  • [next]k8s helm chart for self-hosters past ~100 projects per host
  • [next]desktop dashboard (electron) for offline-edit + git-based workflows
  • [next]briven AI: schema-aware function generator, doc-aware chatbot for self-host operators

not on the roadmap

things people ask about that we don't plan to build, with the why.

  • [no ]swappable storage backend (mongo, planetscale, etc.) — the bet is on postgres. multi-backend is a lot of code for very little user benefit.
  • [no ]edge runtime— deno + cloudflare workers conflict on the network surface; we'd rather invest in regional hosting on real VPSes.
  • [no ]proprietary closed-source plugins — anything we ship lives in the public repo, even if AGPL is incompatible with your use case (we sell exemptions).

miss something? open an issue on code.konnos.org/flndrn/briven.